Testing the Home-Site Advantage in Forest Trees on Disturbed and Undisturbed Sites Eleanor K. O’Brien 1,2 and Siegfried L. Krauss 3,4 Abstract Restoration of plant populations is often undertaken using seed or plants from local sources because it is assumed they will be best adapted to the prevailing conditions. However, the effect of site disturbance on local adaptation has rarely been examined. We assessed local adaptation in three southwestern Australian forest tree species (Eucalyptus marginata, Corymbia calophylla, and Allo- casuarina fraseriana) using reciprocal transplant trials at disturbed and undisturbed sites. Performance of plants within the trials was assessed over 2 years. Planting loca- tion accounted for the majority of the variation in most measures of performance, although significant variation of percent emergence among source populations was also de- tected. In all species, percent emergence and survival of plants sourced from Darling Range populations was sig- nificantly higher than that of plants from the Swan Coastal Plain, regions of contrasting edaphic and climatic environ- ment. Survival of E. marginata over the first 18 months and emergence of C. calophylla were both higher in local plants, providing at least weak evidence for local adapta- tion. Where a local advantage was observed, the relative performance of local and nonlocal seed did not vary among disturbed and undisturbed sites. Evidence for enhanced establishment from local seed in at least one species leads us to recommend that where sufficient high- quality seed supplies exist locally, these should be used in restoration. We also recommend longer-term studies to include the possibility of local adaptation becoming evi- dent at later life history stages. Key words: bauxite mine, genetic variation, home-site advantage, jarrah, local adaptation, Marri, Sheoak. Introduction Local adaptation by natural selection can be a potent evo- lutionary force driving genetic divergence in plant species occupying spatially heterogeneous habitats (Endler 1986; Linhart & Grant 1996). This has been demonstrated experimentally for many species in the form of a ‘‘home site advantage,’’ where plants grown at their site of origin perform better than those translocated to or from more distant sites (e.g., Waser & Price 1985; Galen et al. 1991; Nagy & Rice 1997; Montalvo & Ellstrand 2000; Joshi et al. 2001; McKay et al. 2001; Hufford & Mazer 2003). Adaptive differentiation within species has important consequences for ecological restoration because the intro- duction of maladapted source seed or plants may compro- mise the success of the restored population. Furthermore, interbreeding among individuals from genetically diver- gent source populations may result in genetic swamping or outbreeding depression, causing population decline in subsequent generations (e.g., Fenster & Galloway 2000; Sackville Hamilton 2001; Hufford & Mazer 2003; Potts et al. 2003). In an effort to maintain locally adaptive geno- types and conserve intraspecific genetic diversity, the use of local provenance seed or plants for restoration is com- monly advocated (e.g., Coates & van Leeuwen 1997; Mortlock 2000; Sackville Hamilton 2001; Krauss & Koch 2004). However, strict adherence to such a strategy has been criticized on the grounds that an assumption of local adap- tation may not always be valid (e.g., Lesica & Allendorf 1999; Wilkinson 2001). Sites targeted for restoration are often highly disturbed and may have undergone particu- larly rapid environmental change due to activities such as mining, land clearing, or pollution. Consequently, current selection pressures are likely to differ from those operat- ing historically, and local genotypes may no longer have an advantage (Lesica & Allendorf 1999). It has therefore been argued that the creation of self-sustaining popula- tions on such sites will be better achieved by collecting from a large number of sources to maximize the genetic variation available to selection (Lesica & Allendorf 1999; Wilkinson 2001). From a genetic perspective, the maintenance of locally adapted genotypes in restored populations must be bal- anced against the need to ensure that the population harbors sufficient genetic variation to facilitate adaptive evolution and long-term persistence (Montalvo & Ellstrand 1 School of Animal Biology, University of Western Australia, 35 Stirling Highway Nedlands, Western Australia 6009, Australia 2 Address correspondence to E. K. O’Brien, email eobrien@graduate.uwa. edu.au 3 Kings Park and Botanic Garden, Botanic Gardens and Parks Authority, Fraser Avenue West Perth, Western Australia 6005, Australia 4 School of Plant Biology, University of Western Australia, 35 Stirling Highway Nedlands, Western Australia 6009, Australia Ó 2008 Society for Ecological Restoration International doi: 10.1111/j.1526-100X.2008.00453.x MAY 2010 Restoration Ecology Vol. 18, No. 3, pp. 359–372 359